Closed Bug 1293188 Opened 8 years ago Closed 7 years ago

Crash in EnterBaseline (can be defective hardware [Mem or VGA])

Categories

(Core :: JavaScript Engine, defect, P3)

47 Branch
x86_64
Windows 10
defect

Tracking

()

RESOLVED WORKSFORME

People

(Reporter: bhramcs, Unassigned)

References

(Blocks 2 open bugs)

Details

(Keywords: crash, triage-deferred)

Crash Data

I've been having this crash for a couple of months now, and it occurs like more than once an hour, I'm not sure if it started with a FF update or with a VGA driver update, I'm doubting those 2 factors. This bug was filed from the Socorro interface and is report bp-50064d25-b5c8-469c-8683-fa2b62160807. =============================================================
Component: General → JavaScript Engine
Keywords: crash
Product: Firefox → Core
Flags: needinfo?(jdemooij)
Hi Ahmed, this is almost certainly a hardware problem. I'd recommend running Memtest for at least a few hours. I looked at the crashy JIT code: ecf2dd71: 48 bf d8 66 2a 91 41 movabs $0x241912a66d8,%rdi ecf2dd78: 02 00 00 ... ecf2dd84: 48 bf e8 66 2a 91 41 movabs $0x241912a66e8,%rdi ecf2dd8b: 02 00 00 ... ecf2dd9d: 48 bf f8 66 2a 91 41 movabs $0xff000241912a66f8,%rdi ecf2dda4: 02 00 ff ecf2dda7: 48 8b 3f mov (%rdi),%rdi We crash at the last instruction, because the value we're loading into %rdi is bogus: that 0xff byte should have been 0x00, like the ones before it.
Flags: needinfo?(jdemooij)
Blocks: 1289666
I ran memtest for over an hour, until it completed one pass with no errors. Also I must mention here that no other browser crashes except for FF. Let me know if I can help with any additional info.
I've also noticed in the crash report that it reports 128 TB of virtual memory on my system. I can't imaging how this can be right. This still happens up to 48.01 and very frequently. Again, it appears that only firefox suffers from this problem, there must be a reason for this.
Still a problem on 48.0.2 and still the only browser suffering from this problem is firefox.
Version: 47 Branch → 48 Branch
Have you tried running memory test again? Does crash happen with Firefox started in safe mode?
Flags: needinfo?(bhramcs)
Summary: Crash in EnterBaseline → Crash in EnterBaseline (can be defective hardware [Mem or VGA])
Version: 48 Branch → 47 Branch
It's really good to see someone is still replying to this problem. Here is what's going on so far: 1- Browser doesn't crash now, tabs crash but browser is still open, the browser crash has become very rare, I guess with FF 52. Note: all tabs crash frequently, this is almost the same problem, I have to reload each and every tab again. 2- Can't blame the HW when everything else works fine (AAA games, browsers, videos, ...) only FF has a problem. 3- I tried Safe mode when browser was still crashing (before 52), the crash frequency is certainly less when in safe mode, but it does crash in safe mode. I do love FF, but this problem is really destroying the whole experience.
Flags: needinfo?(bhramcs)
Keywords: triage-deferred
Priority: -- → P3
Your original crash report has aged off the system. Do you have any recent crash IDs?
Flags: needinfo?(bhramcs)
(In reply to Wayne Mery (:wsmwk) from comment #7) > Your original crash report has aged off the system. > Do you have any recent crash IDs? https://crash-stats.mozilla.com/report/index/bp-9fda7415-2125-431c-ba52-469431180102 That's the latest crash report, that's a whole firefox crash. This whole crash has become less and less a couple of releases before quantum, but instead the frequency of tabs crashing is still the same if not more than before. Here is a single tab crash report, happened when I tried to open youtube while typing this response: https://crash-stats.mozilla.com/report/index/bp-abc1abf6-bed7-4389-b2f2-eb90e1180102 A comment: It certainly happens more with a slower connection, I'm not sure if this info is of any help. Best Regards
Flags: needinfo?(bhramcs)
Ignore comment 9, the user commented back that it was indeed bad hardware (memory).
Closing in favor of meta Bug 858032. Issues here seem to be tracked to bad memory.
Status: UNCONFIRMED → RESOLVED
Closed: 7 years ago
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.